forked from ggml-org/llama.cpp
-
Notifications
You must be signed in to change notification settings - Fork 429
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Nix documentation using non-existent config option #1367
Comments
Pinging @DontEatOreo |
Update: Both solutions have now been tested locally and seem to resolve this |
Ping @DontEatOreo |
Sorry for the delay! I've had a bunch of other notifications and this slipped |
DontEatOreo
added a commit
to DontEatOreo/koboldcpp
that referenced
this issue
Mar 2, 2025
DontEatOreo
added a commit
to DontEatOreo/koboldcpp
that referenced
this issue
Mar 2, 2025
DontEatOreo
added a commit
to DontEatOreo/koboldcpp
that referenced
this issue
Mar 2, 2025
LostRuins
pushed a commit
that referenced
this issue
Mar 3, 2025
* nix_example.md: add override example * nix_example.md: drop graphics example, already basic nixos knowledge * nix_example.md: format * nix_example.md: Vulkan is disabled on macOS Disabled in: NixOS/nixpkgs@1ccd253 * nix_examples.md: nixpkgs.config.cuda{Arches -> Capabilities} Fixes: #1367
LostRuins
pushed a commit
that referenced
this issue
Mar 4, 2025
Squashed commit: [b8362da] updated lite [ed6c037] move nsigma into the regular sampler stack [ac5f61c] relative filepath fixed [05fe96a] export template [ed0a5a3] nix_example.md: refactor (#1401) * nix_example.md: add override example * nix_example.md: drop graphics example, already basic nixos knowledge * nix_example.md: format * nix_example.md: Vulkan is disabled on macOS Disabled in: NixOS/nixpkgs@1ccd253 * nix_examples.md: nixpkgs.config.cuda{Arches -> Capabilities} Fixes: #1367 [675c62f] AutoGuess: Phi 4 (mini) (#1402) [4bf5698] phrasing [b8c0df0] Add Rep Pen to Top N Sigma sampler chain (#1397) - place after nsigma and before xtc
LostRuins
added a commit
that referenced
this issue
Mar 4, 2025
…epo size added dpe colab (+8 squashed commit) Squashed commit: [b8362da] updated lite [ed6c037] move nsigma into the regular sampler stack [ac5f61c] relative filepath fixed [05fe96a] export template [ed0a5a3] nix_example.md: refactor (#1401) * nix_example.md: add override example * nix_example.md: drop graphics example, already basic nixos knowledge * nix_example.md: format * nix_example.md: Vulkan is disabled on macOS Disabled in: NixOS/nixpkgs@1ccd253 * nix_examples.md: nixpkgs.config.cuda{Arches -> Capabilities} Fixes: #1367 [675c62f] AutoGuess: Phi 4 (mini) (#1402) [4bf5698] phrasing [b8c0df0] Add Rep Pen to Top N Sigma sampler chain (#1397) - place after nsigma and before xtc (+3 squashed commit) Squashed commit: [87c52b9] disable VMM from HIP [ee8906f] edit description [e85c0e6] Remove Unnecessary Rep Counting (#1394) * stop counting reps * fix range-based initializer * strike that - reverse it
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Describe the Issue
The documentation for how to run KoboldCpp under NixOS uses a non-existent config option, nixpkgs.config.cudaArches
How to reproduce:
Result:
KoboldCpp crashes due to it being compiled with sm_60, instead of sm_75 as specified:
ggml/src/ggml-cuda/template-instances/../fattn-wmma-f16.cuh:422: ERROR: CUDA kernel flash_attn_ext_f16 has no device code compatible with CUDA arch 600. ggml-cuda.cu was compiled for: 600
Additional Information:
GPU: Nvidia GeForce RTX 2070 Super
System Information:
Potential solutions:
koboldcpp.override { cudaArches = ["sm_75"]; }
nixpkgs.config.cudaCapabilities = ["7.5"];
The text was updated successfully, but these errors were encountered: